R/irma_check.R

In Luisiglm/KBoost: Inference of gene regulatory networks from gene expression data

#' Function to produce the AUPR and AUROC Results on the IRMA datasets.
#' @param g a number larger than 0 that is the width parameter for the RBF Kernel
#' @param v a number between 0 and 1 that is the shrinkage parameter
#' @param ite an integer with number of iterations.
#' @export
#' @return list with aurocs and auprs for IRMA datasets
#' @examples
#' res = irma_check()
#'
irma_check <- function(v,g,ite){
    # Pre-allocate memory for the results.
    aurocs <- matrix(0,2,1)
    auprs <- matrix(0,2,1)
    # Run the data.
    grn_on <- kboost(KBoost::irma_on,v = v,g = g, ite = ite)
    grn_off <- kboost(KBoost::irma_off,v = v,g = g, ite = ite)
    # Read the gold standard.
    g_mat = KBoost::IRMA_Gold
    # Now obtain the AUROCS and the AUPRS.
    a <- AUPR_AUROC_matrix(grn_on$GRN,g_mat,FALSE,seq_len(5))
    aurocs[1] <- a$AUROC
    auprs[1] <- a$AUPR
    a <- AUPR_AUROC_matrix(grn_off$GRN,g_mat,FALSE,seq_len(5))
    aurocs[2] <- a$AUROC
    auprs[2] <- a$AUPR
    return(list(aurocs = aurocs, auprs = auprs))
}